Alex Lambert and Liz McKeown to Ed Humpherson: Household finance statistics
Dear Ed,
Household Finance Statistics Quality Improvement Plan
We are writing to update you on progress with the quality improvement of our Household Finance statistics and to set out the user-facing outcomes we intend to deliver over the next three years. Since our previous update, we have made progress in strengthening the quality, coherence, and transparency of our wealth statistics. This includes improved quality and methods information, publication of revised Wealth and Assets Survey (WAS) Round 7 (April 2018 to March 2020) wealth estimates using the updated defined benefit (DB) pension methodology, further work on a consistent historical series, and strengthened governance and stakeholder engagement.
However, while we are making good progress, improvements are not yet complete. In particular, we cannot fully address OSR’s requirement concerning the quality of WAS Round 9 (April 2022 to March 2024) until production and quality assurance have been completed, and users have had an opportunity to consider the results. We therefore do not intend to seek reaccreditation until after the next round of statistics has been published. We will keep users regularly updated on survey performance via the ONS Survey Sample and Response Tracker for which WAS data is now available up to Round 10 (April 2024 to March 2026). Please note, as the tracker is updated periodically, it may not fully reflect the most recent improvements in achieved sample size and sample representativeness.
This letter summarises the progress made, the principal work still required and the outcomes we intend to deliver. The detailed 3-year delivery plan is provided in Annex A. Detailed progress against OSR’s requirements for wealth statistics, including DB pension methodology, are provided in Annex B.
We welcome OSR’s feedback on our plans, and in the meantime, we will continue to engage with stakeholders to communicate any ongoing concerns and developments.
Yours sincerely,
Alex Lambert, Director of Social Surveys
Liz McKeown, Director of Economic Statistics Production and Analysis
Annex 1: Household Finance Improvement Plan: User-facing Outcomes
Year of delivery: 2026/2027
| 1 | Published regular UK household income and family spending statistics, with clear information on how to use and interpret the data given the impact of decreases in sample size |
|---|---|
| 2 | Published back-series of household wealth data in Great Britain, embedding the updated methodology for estimating Defined Benefit pension wealth, enabling users to make direct comparisons over time |
| 3 | Improved suite of quality and methodology guidance for household income and spending statistics, supporting users to analyse data |
| 4 | New formal engagement mechanism introduced for income and spending statistics, following the review of current arrangements - giving our key users clearer routes to provide feedback on our data collection and outputs |
Year of delivery: 2027/2028
| 1 | Family Spending in the UK, FYE 2027 will be published following boost to Living Costs and Food (LCF) Survey issued sample, aiming to deliver 9,000 achieved cases a year. This is likely to first be met in 2027/8 due to ongoing field headcount recruitment to increase overall headcount |
|---|---|
| 2 | Published regular UK household income and family spending statistics with clear information on how to use and interpret the data, including any considerations around survey sample size as this improves |
| 3 | Further improved timeliness of Family Spending in the UK publications, reducing lag between final data collection and publication to be closer to 12 months |
| 4 | Publication of Wealth and Asset Survey (WAS) Round 9 (April 2022 to March 2024) with a full assessment of data quality, alongside an established route towards indexed microdata suitable for record-level linkage to alternative data sources |
| 5 | Production of updated individual and household admin-based income statistics (ABIS) for the first time since 2022, embedding strategic method improvements |
| 6 | Plans for streamlining our income publications communicated to users, setting out what improvements we will implement to help users better navigate our outputs |
Year of delivery: 2028/2029
| 1 | Family Spending in the UK, FYE 2028 will be aligned with current international standards for classifying household expenditure (COICOP18) |
|---|---|
| 2 | Both survey-based income and spending statistics now published within 12 months of collection, improving the timeliness for users |
| 3 | Established route to publishing wealth statistics within 18 months of collection for WAS Round 11 (April 2026 to March 2028) |
Annex 2: Progress update against the review of statistics from the Wealth and Assets Survey
Requirement 1
In recent correspondence, you asked us to ensure that future decisions around DB pension methodology reflect the full range of user needs and consider where ONS could better support users to navigate this complex landscape. Sequencing is key, and in the short-term, we are looking to hold a round-table with key stakeholders to explore the best way of meeting diverse user needs, weighed up against broader priorities for wealth statistics. The outcomes of which will help shape our longer-term work.
Requirement 2
WAS Round 7 data tables and headline release with updated pension methodology applied were published in July 2026, allowing for year-on-year comparisons between Round 7 and Round 8 pension and household total wealth data tables. We also commissioned the Government Actuary’s Department to produce a back-series of pension wealth annuity factors, with WAS back-series scheduled for publication in Autumn 2026, and microdata will be made available on the UK Data Service. We are also producing these data for all previous rounds of the survey, giving a complete and consistent back series, which we will publish this Autumn.
In recent correspondence, you asked us to ensure that future decisions around DB pension methodology reflect the full range of user needs and consider where ONS could better support users to navigate this complex landscape. Sequencing is key, and in the short-term, we are looking to hold a roundtable with key stakeholders to explore the best way of meeting diverse user needs, weighed up against broader priorities for wealth statistics. The outcomes of which will help shape our longer-term work.
Requirement 3
A household finances stakeholder engagement plan was published as part of the December progress update to ONS’ Economic Statistics Plan. The Expert Group will continue playing a pivotal role in shaping future design and methodological changes.
Requirement 4
In March 2026 we published transformed quality documents in line with ONS’ new gold standard approach; this includes a Household total wealth in Great Britain Quality and Methods Guide and a Wealth and Assets Survey Guide; ONS’ first Data Sources Guide as part of ongoing content transformation. These reports are living documents, with mechanisms to incorporate further updates as required. Further, an expanded set of quality indicators has now been published, whereby all main estimates of household wealth have accompanying confidence intervals. The team are further developing tools to produce confidence intervals for distributional wealth estimates, to improve the production efficiency of future WAS outputs moving forwards.
Requirement 5
In December 2025, we published our priorities for improving the quality of wealth statistics. Furthermore, an end-to-end review of resource was used to inform ONS business planning for 2026/27, resulting in small increases in staff resource to priority areas. Onboarding of a Delivery Manager and dedicated Project Management support provides clearer governance and milestone management across parallel WAS rounds, while creation of additional development teams in both surveys and analytical areas better supports necessary development and business as usual activities to run in parallel.
We are maturing production pipelines and reducing dependency on legacy systems, with continued migration to R-based reproducible analytical pipelines for post-collection processing. This includes re-developing and deploying updated outlier checks, consolidating pre-imputation checks and data deliveries for imputation. With a greater emphasis on transparency with users around production processes, we have also hosted walk-through sessions with the Wealth Expert User Group and will continue to share detailed progress.
In parallel, survey operations continue to be on more sustainable footing; with Round 11 now in the field from April 2026 and aiming for an achieved household sample of 18,000 divided equally across the 2 years of the Round. This would bring the issued sample size in line with pre-pandemic levels seen for Round 7 (April 2018 to March 2020).